Documentation |
The Memoryless Nonlinearity block applies a memoryless nonlinearity to a complex, baseband signal. You can use the block to model radio frequency (RF) impairments to a signal at the receiver.
This block accepts a column vector input signal.
The Memoryless Nonlinearity block provides five different methods for modeling the nonlinearity, which you specify by the Method parameter. The options for the Method parameter are
Cubic polynomial
Hyperbolic tangent
Saleh model
Ghorbani model
Rapp model
The block implements these five methods using subsystems underneath the block mask. For each of the first four methods, the nonlinearity subsystem has the same basic structure, as shown in the following figure.
Nonlinearity Subsystem
For the first four methods, each subsystem applies a nonlinearity to the input signal as follows:
Multiply the signal by a gain factor.
Split the complex signal into its magnitude and angle components.
Apply an AM/AM conversion to the magnitude of the signal, according to the selected Method, to produce the magnitude of the output signal.
Apply an AM/PM conversion to the phase of the signal, according to the selected Method, and adds the result to the angle of the signal to produce the angle of the output signal.
Combine the new magnitude and angle components into a complex signal and multiply the result by a gain factor, which is controlled by the Linear gain parameter.
Each subsystem implements the AM/AM and AM/PM conversions differently, according to the Method you specify. The Rapp model does not apply a phase change to the input signal. The nonlinearity subsystem for Rapp model has following structure:
Nonlinearity Subsystem for Rapp Model
The Rapp Subsystem applies nonlinearity as follows:
Multiply the signal by a gain factor.
Split the complex signal into its magnitude and angle components.
Apply an AM/AM conversion to the magnitude of the signal, according to the selected Method, to produce the magnitude of the output signal.
Combine the new magnitude and angle components into a complex signal and multiply the result by a gain factor, which is controlled by the Linear gain parameter.
If you want to see exactly how the Memoryless Nonlinearity block implements the conversions for a specific method, you can view the AM/AM and AM/PM subsystems that implement these conversions as follows:
Right-click on the Memoryless Nonlinearity block and select Mask > Look under mask. This displays the block's configuration underneath the mask. The block contains five subsystems corresponding to the five nonlinearity methods.
Double-click the subsystem for the method you are interested in. This displays the subsystem shown in the preceding figure, Nonlinearity Subsystem.
Double-click on one of the subsystems labeled AM/AM or AM/PM to view how the block implements the conversions.
The following illustration shows the AM/PM behavior for the Cubic polynomial and Hyperbolic tangent methods:
The AM/PM conversion scales linearly with input power value between the lower and upper limits of the input power level (specified by Lower input power limit for AM/PM conversion (dBm) and Upper input power limit for AM/PM conversion (dBm)). Beyond these values, AM/PM conversion is constant at the values corresponding to the lower and upper input power limits, which are zero and $$(\text{AM/PMconversion})\cdot (\text{upperinputpowerlimit}-\text{lowerinputpowerlimit})$$, respectively.
The following figure shows, for the Saleh method, plots of
Output voltage against input voltage for the AM/AM conversion
Output phase against input voltage for the AM/PM conversion
You can see the effect of the Memoryless Nonlinearity block on a signal modulated by 16-ary quadrature amplitude modulation (QAM) in a scatter plot. The constellation for 16-ary QAM without the effect of the Memoryless Nonlinearity block is shown in the following figure:
You can generate a scatter plot of the same signal after it passes through the Memoryless Nonlinearity block, with the Method parameter set to Saleh Model, as shown in the following figure.
This plot is generated by the model described in Illustrate RF Impairments That Distort a Signal with the following parameter settings for the Rectangular QAM Modulator Baseband block:
Normalization method set to Average Power
Average power (watts) set to 1e-2
The following sections discuss parameters specific to the Saleh, Ghorbani, and Rapp models.
The Input scaling (dB) parameter scales the input signal before the nonlinearity is applied. The block multiplies the input signal by the parameter value, converted from decibels to linear units. If you set the parameter to be the inverse of the input signal amplitude, the scaled signal has amplitude normalized to 1.
The AM/AM parameters, alpha and beta, are used to compute the amplitude gain for an input signal using the following function:
$${F}_{AM/AM}(u)=\frac{\text{alpha}*u}{1+\text{beta}*{u}^{2}}$$
where u is the magnitude of the scaled signal.
The AM/PM parameters, alpha and beta, are used to compute the phase change for an input signal using the following function:
$${F}_{AM/PM}(u)=\frac{\text{alpha}*{u}^{2}}{1+\text{beta}*{u}^{2}}$$
where u is the magnitude of the scaled signal. Note that the AM/AM and AM/PM parameters, although similarly named alpha and beta, are distinct.
The Output scaling (dB) parameter scales the output signal similarly.
The Input scaling (dB) parameter scales the input signal before the nonlinearity is applied. The block multiplies the input signal by the parameter value, converted from decibels to linear units. If you set the parameter to be the inverse of the input signal amplitude, the scaled signal has amplitude normalized to 1.
The AM/AM parameters, [x_{1} x_{2} x_{3} x_{4}], are used to compute the amplitude gain for an input signal using the following function:
$${F}_{AM/AM}(u)=\frac{{x}_{1}{u}^{{x}_{2}}}{1+{x}_{3}{u}^{{x}_{2}}}+{x}_{4}u$$
where u is the magnitude of the scaled signal.
The AM/PM parameters, [y_{1} y_{2} y_{3} y_{4}], are used to compute the phase change for an input signal using the following function:
$${F}_{AM/PM}(u)=\frac{{y}_{1}{u}^{{y}_{2}}}{1+{y}_{3}{u}^{{y}_{2}}}+{y}_{4}u$$
where u is the magnitude of the scaled signal.
The Output scaling (dB) parameter scales the output signal similarly.
The Linear gain (dB) parameter scales the input signal before the nonlinearity is applied. The block multiplies the input signal by the parameter value, converted from decibels to linear units. If you set the parameter to be the inverse of the input signal amplitude, the scaled signal has amplitude normalized to 1.
The Smoothness factor and Output saturation level parameters are used to compute the amplitude gain for the input signal:
$${F}_{AM/AM}(u)=\frac{u}{{\left(1+{\left(\frac{u}{{O}_{sat}}\right)}^{2S}\right)}^{1/2S}}$$
where u is the magnitude of the scaled signal, S is the Smoothness factor, and O_{sat} is the Output saturation level.
The Rapp model does not apply a phase change to the input signal.
The Output saturation level parameter limits the output signal level.
The following describes specific parameters for each method.
Scalar specifying the linear gain for the output function.
Scalar specifying the third order intercept.
Scaler specifying the AM/PM conversion in degrees per decibel.
Scalar specifying the minimum input power for which AM/PM conversion scales linearly with input power value. Below this value, the phase shift resulting from AM/PM conversion is zero.
Scalar specifying the maximum input power for which AM/PM conversion scales linearly with input power value. Above this value, the phase shift resulting from AM/PM conversion is constant. The value of this maximum shift is given by:
$$(\text{AM/PMconversion})\cdot (\text{upperinputpowerlimit}-\text{lowerinputpowerlimit})$$
Scalar specifying the linear gain for the output function.
Scalar specifying the third order intercept.
Scalar specifying the AM/PM conversion in degrees per decibel.
Scalar specifying the minimum input power for which AM/PM conversion scales linearly with input power value. Below this value, the phase shift resulting from AM/PM conversion is zero.
Scalar specifying the maximum input power for which AM/PM conversion scales linearly with input power value. Above this value, the phase shift resulting from AM/PM conversion is constant. The value of this maximum shift is given by:
$$(\text{AM/PMconversion})\cdot (\text{upperinputpowerlimit}-\text{lowerinputpowerlimit})$$
Number that scales the input signal level.
Vector specifying the AM/AM parameters.
Vector specifying the AM/PM parameters.
Number that scales the output signal level.
Number that scales the input signal level.
Vector specifying the AM/AM parameters.
Vector specifying the AM/PM parameters.
Number that scales the output signal level.
[1] Saleh, A.A.M., "Frequency-independent and frequency-dependent nonlinear models of TWT amplifiers," IEEE Trans. Communications, vol. COM-29, pp.1715-1720, November 1981.
[2] A. Ghorbani, and M. Sheikhan, "The effect of Solid State Power Amplifiers (SSPAs) Nonlinearities on MPSK and M-QAM Signal Transmission", Sixth Int'l Conference on Digital Processing of Signals in Comm., 1991, pp. 193-197.
[3] C. Rapp, "Effects of HPA-Nonlinearity on a 4-DPSK/OFDM-Signal for a Digitial Sound Broadcasting System", in Proceedings of the Second European Conference on Satellite Communications, Liege, Belgium, Oct. 22-24, 1991, pp. 179-184.